08+09 - Heat and Mass Transfer - in 2D Flashcards
1
Q
Heat conduction
A
- Fourier’s law: q = -k.A.dT/dx
- heat flux = q’’ = q/A
- 1D steady state heat conduction: d/dx(k.A.dT/dx) +Q.A = 0
- conduction is b/ 2 different temperatures
2
Q
Essential / Natural BCs
A
- essential: where the value is specified
- natural: that will be verified automatically once the solution is found
3
Q
Convection
A
- q_in = h.A(T_inf - T)
- heat flow through fins
- b/ volume and surface of different T°
4
Q
Mass transfer
A
- J = -D.dC/dx
- dC/dt = D(∂2C/∂x2 + ∂2C/∂y2 +∂2C/∂z2)
- with advection: Gamma = -D.nabla(C) + w.c.U
5
Q
Possible BCs in heat transfer
A
- constant T°
- cstt surface heat flux
- adiabatic surface
- convection surface cdt°
6
Q
FDM for steady-state problems
A
- procedure: discretize -> energy balance method -> solving for unkn. nodal T° -> use T° field and Fourier’s law to determine heat transfer in the medium
- solution methods
- – matrix inversion: T = A-1.C (A coeff matrix, C vector of cstts)
- – Gauss-Siedel iteration: reorg to have diagonal e/ > other row e/, iterate until |T(k,i) - T(k-1,i)| < epsilon (for implicit method)
7
Q
FDM for transient problems
A
- procedure: same except solve until steady-state is reached
- eq°: 1/alpha . ∂T/∂t = nabla2(T) with alpha = k/rho.c_p thermal diffusivity
- methods
— explicit: evaluate at p+1 with T° values from previous time step p
time derivative is a forward difference
stability criterion: delta(t) <= delta(x)2 /4.alpha <=> Fo <= 1/4 = FD form of Fourier number
quicker but conditionally stable
— implicit: evaluate each term at p+1
time derivative is a backward difference
nodal eq° must be solved simultaneously
unconditionally stable but longer